Archives for the month of: September, 2010

You may have noticed that those Brits seem to love to talk about trim sizes in a totally weird way.  Rather than saying 5½” x 8½” they will come out with the outrageous Demy Octavo (which, even worse, they may write Demy 8vo). Note also that if you ever do pin them down to precise measurements they will talk in millimeters and give the dimensions the opposite way round than we do.  Thus a 6” x 9” book will turn into a 228 x 152mm one — if you’re lucky enough not to get bogged down in Medium 8vo.  This back-to-front dimension problem simply has to be accepted as part of the scenery, just as in the old days was the fact that they would always want was negs with emulsion down and we’d want them emulsion up (or was it the other way round?).  Knowing all this does help in dealing with Chinese printers – Hong Kong, Empire, it all makes sense.

The Demy 8vo story goes back to the early days of printing, and relates to the standard sheet sizes used 300 or 400 years ago.  So starting with a demy sheet, which measured 22½” x 17½”, folding it once will give you demy folio; fold it in half again and you’ve got demy quarto, again and it’s demy octavo, 8¾” x 5-5/8”.  Trim 1/8” off top & bottom and the fore-edge and you have a finished book trim size 5½” x 8½”.  Of course, even in the world of sheet-fed printing almost nobody has printed on a demy sheet for aeons: but in the middle of the last century Double and Quad Demy were not uncommon.

Here are a few of the old sheet/press sizes:

Double Quad Crown:  60” x 40”

Quad Demy:  45” x 35”

Super Royal: 27½” x 20½”

Royal: 25” x 20”

Medium: 23” x 18”

Demy: 22½” x 17½”

Music Demy: 20” x 15½”

Crown: 20” x 15”

Those who are unfamiliar with this printing process will probably think I’m talking nonsense when I say collotype is a process for printing in continuous tone without the use of a halftone screen (ie. NO dot).  Collotype also permits the use of highly pigmented inks so that an intensity of color can be achieved which is unimaginable in the world of four-color process.  Unfortunately the process is no longer commercially available, though there are one or two places where it is still done on an art basis.  In the early 70s I was fortunate enough to supervise the printing of what must have been one of the last few books to be printed by this process.  It was a black & white corpus of Roman coins, and the result was certainly impressive; but worth the expense?

One thing people today don’t realize I suspect is how much art there used to be in getting a book printed.  Straight text was obviously easier, but even that required considerable skill in stroking a rather crude piece of machinery so that it would perform consistently.  When illustrations were involved, especially color, there was an almost miraculous quality to the ability of pressman and other skilled craftsmen to deliver the results they achieved.  Over the past thirty or forty years most of that craft skill has been moved from the craftsman’s brain to the hardware and software of the printing equipment, and nowadays we expect printing to conform to our standards almost automatically.  The “craft” involved in letterpress, offset lithographic, or gravure printing was however as nothing compared to the demands of collotype.  In crude folk terms the process involves coating a sheet of glass with egg white and gelatin, drying it in the sun, and using this plate to print using the same grease/water antipathy as offset lithography does.

The word collotype derives from the Greek, kolla, glue.  “Glue type” doesn’t really give one much understanding of what’s going on when we print by collotype, though it does sound quite fun.  Think of traditional offset, with a negative, and you’ve got most of it.  The magic resides in the plate making.  A glass (or aluminum) plate is used, coated with albumen, gelatin or stale beer!  This is to allow the printing substrate of warm gelatin sensitized with either potassium or ammonium bichromate to adhere firmly to the plate.  After coating, the plate is dried in an oven (formerly in the sun): the oven temperature is critical: too hot and the grain becomes coarse causing too much contrast; too cool and the fine grain will cause the image to wash out.  The image, carried on a negative is them exposed onto the plate.  The bichromated gelatin reacts to the amount of UV light, hardening in proportion to the tones of the image.  The hardening makes the gelatin least absorbent in the tones receiving the most light and most absorbent in the areas receiving little or no light.  The bichromate is then washed from the plate leaving only the gelatin.  Just before printing, the plate is covered with a solution of glycerin and water which is absorbed in proportion to the tones in the image: the shadows absorb least, and the highlights absorb most.  Just as on an offset press the wet areas repel the greasy ink and as a result a continuous tone image is printed on the sheet.  The life of the plate is limited, so long runs are not possible.  The maximum might be 2000 depending on a whole range of variables.

An interesting website is where a British research project involving digital printing with collotype quality is reported.  They really can’t mean “collotype” they just mean “continuous tone” surely.

Everyone (in the book trade) knows that a cast off is a calculation of how many book pages a manuscript will make.  Benjamin Franklin tells us in 1784 “The compositors in your chapel* do not cast off their copy well”.  To people who don’t work with books the meaning is unlikely to be quite so obvious.

The Oxford English Dictionary defines to cast off as “To count or reckon, so as to ascertain the sum of various numbers, orig. by means of counters, to the manipulation of which the word probably refers.”  Think abacus — not a tool used today by most estimators.  Originally the verb “cast” could stand on its own with this meaning: to cast a column of figures, or to cast accounts, but now it seems to need the support of the “off”.  The cast off is one of these wonderful numbers that comes with its own guaranteed margin of error – the margin determined by the skill of the caster-offer.  But it is a number that’s fundamental to all our activities, and severe retribution can result when it’s too far off target.

An analogous usage is to cast a horoscope – basically a calculation – also coming of course with its own built-in margin of error.


Cast and throw are almost synonyms.  Thus we have in about 1400 “Palomydon cast off his clothis cantly & well.” And in 1697 Sir William Dampier refers to a quick acting tree “In a weeks time the Tree casts off her old Robes.”  Milton appears to be the first to say “to cast off this yoke” in Paradise Lost, implying a revolutionary rejection of authority.  Of closer personal impact perhaps we have the same rejection in the King James Version of the Bible (1611) in Psalm lxxi. 9 “Cast me not off in the time of old age.”

The worlds of hawking and hunting give us cast off meaning to unleash (dogs), to let fly (hawks).  For example: “When a magpie is seen at a distance, a hawk is immediately to be cast off.”

Alchemy comes into the picture too, where cast off means to throw off, as for instance vapor, or to run off as melted metal.

Sailing and boating have two different usages of cast off: obviously setting off, untying the rope holding the boat to the dock (basically the same meaning as the hunting examples) and shortening sail as in  “The gaskets cast off the fore topsail”.

In country dancing to cast off means to maneuver around other dancers to a position lower down the set.  In knitting it implies completion of the job: “To cast off which is done by knitting two loops and pulling the first made loop over the last” 1880.

* chapel is an interesting usage in itself.  Printing union ‘branches’ are still referred to as chapels, with the shop steward designated as the father of chapel.

A colophon was originally “the inscription or device, sometimes pictorial or emblematic, formerly placed at the end of a book or manuscript, and containing the title, the scribe’s or printer’s name, date and place of printing” (Oxford English Dictionary).  A modern day example may be found in any Knopf book, where there’s also usually a description and history of the typeface used.

The word has tended to become synonymous with logo (deriving from logogram or logotype) standing for imprint: the devise and standard group of words identifying the publisher on the title page: in the case of Knopf again, the borzoi.

The word logotype goes back to hot metal days when each character was a separate piece of metal.  A logotype was originally a single piece of type metal containing a number of letters or a word.  Obviously if you tended always to put the same ID on your title page it saved time to set it up as a single piece which you’d just pull down whenever needed.  Strangely, a rigid application of this definition makes every line of Linotype officially a logotype.

Logogram refers back to the time of the scribes, who in order to save time and space used a number of logograms, the et of the Latin “and” being a prominent survivor in the shape of the ampersand.  Another example, @ for at, has recently gained a currency way beyond its earlier usage.  Other ligatures set on single piece of metal in hot metal days, and thus definable as logotypes, include ff, st, vr (for very) and the diphthongs ae, oe.

Why would payments to authors make this reference to the Queen?

In the 15th and 16th centuries the word royalty meant a grant made by the monarch – thus for example the Oxford English Dictionary tells us “The Maior and Burgesses of Oxon do stande so muche..uppon theire right and royaltie of the Thames.”  Obviously the king had allowed them to take fish from the river, or row on it, or more probably to charge tolls to people who did those things, since politicians have always loved to do that sort of money raising thing.  Royalties then came to mean payments made to the landowner of a mine for the privilege of extracting the coal from it – no doubt because the monarch was a large landowner, and the word got broadened out to cover quite unroyal folks.  Obviously it was then but a small step when by the middle of the 19th century it was first applied to payments made to authors and to the holders of patents, even though by now Queen Victoria didn’t actually have the right to grant them this income.

Perhaps it adoption in America was sufficiently late for the anti-monarchist ideology of the early Republic to fail to register with those authors and publishers who brought the word over the ocean.  The French find no need to refer to royalty in this context: their word is redevance: I wonder if they did before the Revolution.  Apparently royaute is accepted as meaning royalty in our sense in Quebec only.  Fascinatingly, it appears to be le mot juste in Russian too.

There is no absolute standard for the margins of a book; it’s really a matter of aesthetic balance between the dark type area and the white surround.  Bear in mind that when we look at a page we also see the page facing it, and that has to come into the judgment of balance.  In general we can say that “correct” page layout will end up with the gutter margin being the smallest, the head margin next, the fore-edge next, and the foot margin largest of all.  This enables the reader, in theory, to hold the book with a thumb at the bottom of the page without obscuring any of the text.  Of course de luxe margins, like de luxe anything else in book manufacturing, are now a thing of the past.  But still the proportions should be maintained if possible as long as we care at all about the look and feel of the book.  For those who crave a formula a relationship of 1½ : 2 : 3 : 4  is good old conservative guide.

It’s probably safe to say there are no circumstances under which the gutter margin should be less than ½”.  So with a foot margin approaching 1½” you’re probably going to be told you’re wasting paper if you persevere with this formula.

This one is an example of American terminology being more old-fashioned and quaint than British: an unusual situation, in this business at least.  In Britain the weight of paper is given as grams per square meter (gsm) which is a usable, simple number.

Over here, basis weight is the weight in pounds of a ream of paper cut to a given standard size for that type of paper.  For book paper the standard size is 25” x 38”.  A ream is defined by the OED as “a quantity of paper, properly 20 quires or 480 sheets, but frequently 500 or more, to allow for waste; of paper for printing, 21½ quires or 516 sheets (a printers’ ream)”.  Let’s not worry about quires and printer’s reams: we down-to-earth Americans define the ream as 500 sheets, so a 50# paper (# standing for basis weight) will clock in at 50lbs if you put 500 sheets, size 25″ x 38″ on a scale.  This is of course not very useful information: who has 500 sheets in that strange size lying around?

It all goes back to ancient times, at least 100 years ago, when many American printing presses would accommodate a 25” x 38” sheet, yielding a 6⅛” x 9¼” book. Because common, this size became the standard.

Basis weight is relevant merely because that’s how paper is sold.  If you want to make any use of the number, you really need the M weight, which tells you how much 1000 sheets in the size you are using will weigh.  This enables you to work out how many pounds of paper you’ll need.  M weight is calculated as follows:

M weight = Length of sheet x Width of sheet x Basis weight divided by 475.

Of course this is only interesting if you ever do sheetfed printing.  For those who deal overseas you can convert from Basis Weight to gsm by dividing the Basis Weight by 0.67565.  To go the other way, multiply the gsm by the same magic number.  But do remember these formulae apply only to book papers: other types of papers have different basis weights.

In the beginning printers aimed to make their work as similar as possible to the hand-written books that had preceded them.  As there are today, there were conventional sizes for different types of books: for example, brevier comes from breviary and became the name for a type size (approximating to 8 point).  A pica was an ecclesiastical directory giving the rules for fixing the dates of moveable church feasts.  It perhaps has no relevance to the history of type, but pica is also the Latin for magpie, and the OED speculates that there may have been some etymological connection between the bird and the book.  Obviously directories of moveable feasts must have been of such a size that to fit all the information onto the page you had to write it in script which was about as large as today’s 12pt type.

Like so much in the world of printing and publishing the exact size of a pica has fluctuated with time and location.  A pica is 12 points: that much has remained constant.  The size of a point is really the variable: in the age of computers it is now defined as 1/72”, so the pica is 1/6”.

Traces of the old names for type sizes: bourgeois (pronounced bur-joice like rejoice), minion, canon, primer (pron. primmer), and others named after precious stones; pearl, agate, ruby, can still be found today in the names publishers give to their different sizes of bibles.

The sizes given in the list below are approximate.  When fonts were thus named different printer would cast types to various sizes.  Overall the sizes thus named conformed roughly to the point sizes shown.

Excelsior         3pt

Brilliant           4pt

Diamond         4½ pt

Pearl                5pt

Agate               5½ pt

Ruby               5½ pt

Nonpareil       6pt

Emerald          6pt

Minion            7pt

Brevier             8pt

Bourgeois       9pt

Long Primer  10pt

Small Pica     11pt

Pica                 12pt

English            13pt

Great Primer  18pt

Paragon          20pt

This all goes back to the days of hot metal typesetting and letterpress printing.  A galley was originally a long metal tray into which the lines of hot-metal type were placed after composition.  After the galley was filled with lines of type (about 2-3 pages worth depending on the page design) it would be locked up, holding the type tight within it, then inked by hand roller and a proof pulled.  This galley proof, the first proof, was commonly referred to as a galley.  It would be followed after correction, by page proofs – the galley was unlocked and the lines of type corrected and made up into pages, before a similar lock-up and hand proofing of the composed pages.

A galley of type coming from the caster. Photo: Gloucester Typesetting

A galley of type coming from the caster. Photo: Gloucester Typesetting

So galleys over the years came to mean first proofs, and as publishers began to economize by not sending revised proof to their authors, checking them in-house against the foul galleys, the word began to be synonymous with “proofs”.   Being a more romantic sounding word than just “proof”, “galley” has stuck, but has now morphed to a more specialized meaning as bound galleys – those lovely paperback books we all love to pick up at BEA.

Bookland is an imaginary place invented in the 1980s to provide a “country code” for the EAN which the book trade had decided to move to (expanding on the ISBN system started in UK in 1966 and USA in 1968 – yes I started in this business before there were SBNs). All EANs (European Article Numbers) start with a 3 digit country identifier, indicating where the manufacturer of the product is registered. Because the same book can be available in many countries, and may be manufactured at different (or indeed the same) times in different countries, the original country identifier was not workable for books. Oxford University Press is a perfect example of this simultaneous availability of the same item in many different countries. The solution was to come up with an imaginary country, Bookland, which would be the place of origin of all books. It’s a bit like Santa, actually.

Following after the 978 which identifies Bookland, the ISBN is made up of the following elements. A language indicator: 0 and 1 both mean English. Of course it’s not the language of the book itself that’s indicated, it’s the language of the publisher’s home base. Thus an OUP book in Spanish would still be 978-0-19. In fact the “language indicator” is more accurately a country indicator: for instance Spain is 84 and Mexico is 968 and 970. Next comes the publisher prefix, which will be of varying length depending on the number of ISBNs expected to be issued by that publisher. HarperCollins is 02, Oxford 19, Cambridge University Press 521, New York Review Books 59017 and so on. The remaining digits, except the last, are unique identifiers of the book, with the last one being a check digit.

The check digit is there so that any computer set up to deal with books can identify whether the numbers it has just been fed are a valid ISBN or whether there’s a typo in the numbers. The calculation methods are different for ISBN-10 and ISBN-13, though they both involve calculating the “remainder” after a division. For an ISBN-10 the first nine digits in order are multiplied by 1, 2, 3, 4, 5, 6, 7, 8, and 9, then the resulting answers are added up. That total is divided by 11, and the remainder is the check digit. If it’s 10, then the check digit becomes an X.

Calculation of the ISBN-13 check digit is done using a modulus 10 formula.
Each of the first 12 digits is alternately multiplied by 1 or 3. The check digit is equal to 10 minus the remainder resulting from dividing the sum of the weighted products of the first 12 digits by 10. The formula is mod 10 (10 – mod 10[X]) = Check Digit.

An example:

9 7 8 0 1 9 5 3 1 1 4 5 ?
x 1 x 3 x 1 x 3 x 1 x 3 x 1 x 3 x 1 x 3 x 1 x 3
9 + 21+ 8 + 0 + 1 + 27+ 5 + 9 + 1 + 3 + 4 + 15 + = 113

113/10 = 11 remainder 3

10-3 = 7. So check digit is 7.

Think of the EAN barcode as the numbers of the ISBN set in a different typeface, a typeface consisting of thick and thin stripes. To the right of the barcode is (usually) a series of shorter bars which are for the price. A book’s price tends to be more changeable than many products’ (or the inventory tends to sell over more price cycles than most products) so publishers tend to leave this price add-on empty: it appears as 90000 in this case. You may see the price add-on in use on faster moving titles from trade publishers.

After initial resistance by the less numerate members of the publishing community (about 99.99% of them in the sixties) ISBNs have obviously caught on. It’s sometimes a little frustrating when operations staff refer only to ISBNs, as if the objects we deal in didn’t have any other identifying features, and as if customers were purchasing a lottery number rather than a book with interesting contents.